Air Out, Energy Efficieny In
- By Ralph Morrone
- April 1st, 2009
Youngstown State University (YSU), located in northeastern Ohio, encompasses approximately 4M sq. ft. Approximately 2M sq. ft. of the gross square footage is occupied as usable space. In 2006, YSU entered into a Performance Contract (PC) with a building efficiency and power solutions company to save the University nearly $13M through 10 years. These reductions in energy were accomplished via T12-T8 lighting conversions, steam trap surveys, control enhancements, Variable Frequency Drive (VFD) implementations, and replacement of existing chillers with high-efficiency chillers.
While the project as a whole is serving the University well, YSU sought to better optimize the new chiller plant in order to achieve more savings. The former plant included three 1,200-ton chillers and a 1.3M gal. Thermal Energy Storage (TES) tank that worked to meet the campus’s peak loading of 3,500 tons. The retrofit converted a primary/variable secondary pumping system to a variable primary pumping system. Following the retrofits and efficiency measures, the campus was calculated to have a loading of approximately 3,000 tons on a degree day.
In 2007, the first year in operation, YSU’s plant appeared to have a maximum chiller output of roughly 2,400 tons. This output was based on plotted data of hourly enthalpy readings vs. tonnage. See 2007 Chiller Output trend line and data points (RED
) in Figure 1
On a 2,400-ton peak output day, YSU observed a temperature difference across the chillers of 8.5°F. Because of inherent system flow and pressure issues, which are currently being resolved, system flow to campus requires 6,500 GPM. Because flow and pressure requirements are dictated by individual buildings, the central plant’s tonnage was determined by the chilled water coils’ ability to transfer heat. This transfer of heat directly affected the Delta T across campus since flow was held constant.
Tonnage is expressed by the following equation:
Or in our case:
YSU, upon learning of a technical briefing, was intrigued by discussions illustrating the differences between entrapped, entrained, and dissolved air and the theoretical efficiencies attained by removing all three from a hydronic system. The catch to complete air separation though was to utilize a coalescing air separator versus a conventional air separator.
is air that is traveling in the water stream in bubbles that can be seen with the human eye, while entrapped air
is air that has migrated out of the water stream and into high points as pockets of air. Dissolved air
are microscopic bubbles that cannot be seen in a water stream but exist in levels measured in parts per million and coat heat exchangers as the velocity of the water is slowed down through any coils. The prevalent and conventionally used air removal devices will relieve entrained air when sized properly but are not designed to affect entrapped or dissolved air. YSU was currently utilizing a conventional air separator.
A leading manufacturer of coalescing type separators came to YSU and utilized a portable demonstration which illustrated their claim as to how a super-efficient coalescing air separator could rid the system of up to 99.6 percent of system air. The separator starts with removing the entrained air as the water enters the separator. Once the entrained air was exercised from the system, the dissolved air content was then taken out of the water with similar science. Once the dissolved air is removed from the system, the water will actually pull the entrapped air into the stream of water, and it will then be taken out by the separator, thus eliminating up to 99.6 percent of all air in the system. These last two steps, we were told, are not usually completed with most conventional air separators.
After some careful consideration and a desire to increase efficiency, YSU made the commitment to test and document the theory behind this claim to air separation superiority. While the initial thought was to test one hydronic heating system, since the efficiency gains on heating systems are much larger, YSU opted to treat the campus chilled water loop instead. After installing a 20-in. coalescing separator rated for total plant pump capacity in 2008, post installation data (BLUE
) was compared to the previous year (RED
). Those results are listed below in Figure 2
This data was compiled via humidity and temperature readings from our central plant, along with the tonnage readings from our chillers. The humidity and temperature readings were interpolated to enthalpy readings. This enthalpy and tonnage were then plotted for the entire summer of 2007 (pre coalescing separator), and the end of summer 2008 (post coalescing separator). By utilizing enthalpy, temperature variations, along with humidity swings, could still allow for like comparisons between different years with all else being held equally.
While YSU expected that the system enhancements would achieve efficiency and scaled-back tonnage out to campus, the exact opposite was seen. Upon further review, YSU realized total tonnage out to campus averaged 15.9 percent more than what we were able to attain with the previous air separator under the same operating parameters in 2007. This increase in tonnage was a direct result of the inability of chilled water coils to absorb heat due to system air not being removed with conventional air separators. Because chilled water flow to campus is presently fixed, the plant’s output capacity is at the mercy of the delta T across campus. Without a doubt, the coalescing air separator had a quantitative and positive impact on YSU’s chilled water system and chiller tonnage output.
Additionally in 2007, on summer design days, YSU saw outlying buildings with the longest equivalent pipe length from the central plant struggling to come within 5°F of Discharge Air Temperature (DAT) set-points. As a result of seeing a higher tonnage output to campus, the Delta T across-campus coils met their design intents, and two-way valves were beginning to close when previously they would have still been stroked open. This cause and effect allowed these closed valves to force chilled water flow to the extremities, and ultimately cool ailing air handlers and DATs everywhere on campus.
Another inherent benefit to the air removal and higher Delta T on the chilled water supply and return was the efficiency gains on the chillers. The new York electric centrifugal chillers were optimized for a 10°F Delta. By allowing the chillers to operate under design conditions, YSU is able to recognize the maximum efficiency attainable by the equipment. Further efficiency gains will also be recognized as YSU addresses the flow constraints dictated to the plant by the buildings on campus. As the chilled water system’s efficiency gains translate into energy savings at the chillers, additional savings will be gained when YSU is able to take advantage of its variable primary pumping with reduced pump energy needed to deliver energy to buildings.
Some planned additional energy measures to further reduce campus energy are campus-wide air-side coil cleaning and revised preventative maintenance schedules. As YSU endeavors towards an ultimate goal of achieving minimum kW/ton, it is known that the once the chilled water system is optimized, the hydronic heating systems are next. YSU looks forward to implementing the same proven energy efficiency measures on the heating systems in each building. As previously stated, the chilled water system’s anticipated energy savings should translate into even more savings with the heating systems on campus.
Ralph Morrone is a facilities engineer for Youngstown State University. He can be contacted via e-mail at firstname.lastname@example.org.